Paperclip 📎 gives your agent access to millions of full-text biomedical papers.
Agents like Claude Code, Cursor, and Codex are powerful research environments, not just for coding. But searching for papers often returns missing or incomplete results because these agents only have access to basic web search.
With our release of Sy (https://t.co/hZK8RxcPCi), we showed that giving agents access to a deep index of biomedical papers improves literature search (1.6× more accurate, 2.4× faster, and 3.6× cheaper) over Claude Code alone.
🚀 Today, we're releasing Paperclip as a CLI that can be installed directly in your agent IDE!
👉 To install and try it out, go to https://t.co/tGSkYhbrZx
📖 We also wrote a technical deep-dive into how Paperclip works and why agents take to it really well: https://t.co/5iAJ6zv58r
Excited to announce re:AGENT a two-day build weekend (8/15–8/16, SF) for people building tools for scientific agents.
Three tracks:
A) Build an AI scientist that can carry out a workflow from start to finish
B) Design a new biological system from the ground up
C) Create a dataset at scale to answer a question you've never been able to read fast enough to settle.
What you leave with is the working thing, not a pitch deck.
Apply by 7/30 at https://t.co/Wo0QrFkZ6J
https://t.co/GGp4yCV9pB is taking over SF in August for re:AGENT - a hackathon for end-to-end agentic science
brought to you by some of the goats of bio ai
> @AnthropicAI
> @arcinstitute
> @gxl_ai
> @biohub
> @benchflow_ai
you know this one's gonna be good
link below
Bio databases/APIs weren't designed for AI agents🧬
So we expanded #Paperclip to transform common molecular databases—UniProt, PDB, ChEMBL—into a unified agent-native virtual filesystem.
Now you can do deep research across 10^6 molecules much more accurately and 10x faster!
All integrated with millions of full-text papers also indexed in Paperclip. Free to use https://t.co/1b8XZR7KuV
Paperclip is now in your browser.
Today we're releasing Paperclip Chat, an instant, browser-based way to search and analyze millions of full-text papers, clinical trials and regulatory documents with no local setup.
Paperclip contains every feature of the initial CLI, plus:
- Inline citations
- Repo visualization: see your research as a graph: click through each branch, inspect verified claims, full observability into every conclusion
- A workspace where you can upload your own documents
Try it → https://t.co/Ss030Q3nNB
More → https://t.co/KNuCVikpbY
Paperclip original CLI → https://t.co/Mw8Y7foT8G
Excited to introduce Paperclip Chat!🖇️
Now you can use #Paperclip directly in your browser for deep literature synthesis for free
⚡️ Each claim is grounded with inline citations
⚡️ Upload your own PDFs. They're automatically integrated into Paperclip's virtual filesystem
@anshulkundaje Can’t just tout the papers that are positive and critique the papers that are negative. Same problems are present in the papers they like.
They decline to participate in https://t.co/i06dY1fFsY and similar. Blinded head to head comparison >> benchmarks. @ericwu93@kevinywu
Today, #paperclip is even faster and more reliable.
We introduce Paperclip Repos, a git-like layer on our virtual filesystem that makes literature synthesis reproducible + verifiable.
This enables better grounding and large-scale scientific analysis https://t.co/HjJUqnb2rC
Paperclip is the only "what does scientific literature say about XYZ?" skill I've tried that doesn't suck.
Already had all of Pubmed + {bio,med}rxiv + 150 million OpenAlex abstracts...
This is icing on the cake 😍
We added >220K FDA regulatory and >1M clinical trial docs to #paperclip. All natively indexed for agents and free.
Now agents can easily reason over clinical studies w/o web search!
E.g: find all trials that were approved despite missing endpoint https://t.co/30GGqfCQmO
@LH@james_y_zou@gxl_ai Yes! Paperclip creates a Claude Code skill file (if selected) when it is installed. Alternatively, you can use it via MCP. https://t.co/WTOiAJ79BS
@Chain_AlphaX@james_y_zou@gxl_ai Not at the tool call level, but overall agent execution. We observe that default Claude Code websearch + webfetch often has to make many sequential calls to answer a question, whereas paperclip grep + map provides very dramatic speedup by performing parallel pre-indexed searches.
@girishkaitholil@james_y_zou@gxl_ai Paperclip doesn't perform deduplication by default, meaning both versions will show in the search results. We found in practice, AI agents are quite good at figuring out which versions are more recent/relevant based on the available metadata (date, journal, etc.).
@therealRYC@james_y_zou@gxl_ai Hi Robert, Paperclip updates automatically with the latest literature as its made available, depending on the source (e.g., biorxiv makes monthly releases).
Big Update🤩: #paperclip now includes full papers from all of arXiv, PubMed Central and 150 million abstracts!🖇️
You can give your LLM all that knowledge in one line—all optimally indexed for AI agents. Much more thorough and ~100x faster than web search, and free.
Big Update🤩: #paperclip now includes full papers from all of arXiv, PubMed Central and 150 million abstracts!🖇️
You can give your LLM all that knowledge in one line—all optimally indexed for AI agents. Much more thorough and ~100x faster than web search, and free.